Re: Firefox HTTPS-only mode breaks sites that return 404 for HTTPS connections

[Kenneth Parker]

On Wed, Apr 14, 2021, 11:57 PM Joe Pfeiffer  wrote:

> Kenneth Parker  writes:
>
> > On Wed, Apr 14, 2021 at 10:16 PM Celejar  wrote:
> >
> >  On Wed, 14 Apr 2021 22:57:01 +0100
> >  piorunz  wrote:
> >
> >  > On 14/04/2021 17:19, Celejar wrote:
> >  > > Hi,
> >  > >
> >  > > I recently switched to Firefox's native HTTPS-Only mode from the
> >  > > HTTPS Everywhere extension, and I've just made the nasty discovery
> that
> >  > > a bunch of links that had been returning 404, which I had been
> assuming
> >  > > were dead links, were actually perfectly valid pages, which Firefox
> had
> >  > > been "upgrading" to HTTPS, and then getting 404s from web servers
> that
> >  > > weren't offering them via HTTPS (but still apparently accepting
> >  > > connections via HTTPS). Clicking on the little lock icon and turning
> >  > > HTTPS-Only mode off for the website doesn't seem to have any effect
> -
> >  > > the only thing that lets me actually access these pages is turning
> off
> >  > > HTTPS-Only mode via the general Settings page (or about:config).
> >  > >
> >  > > When the website doesn't offer HTTPS at all, then Firefox offers to
> >  > > connect via HTTP, after a warning, and that's fine. But having pages
> >  > > become completely inaccessible is intolerable - I now have to check
> >  > > every 404 I get by turning off HTTPS-Only mode and seeing if the
> >  > > page is actually there. Am I missing something here, or is
> HTTPS-Only
> >  > > mode just badly broken?
> >  >
> >  > It certainly works fine for me. I use https only mode for many months
> >  > now. Can you bring an example of a page which returns good page on
> http,
> >  > but 404 error on https?
> >
> >  http://www.daat.ac.il/
> >  https://www.daat.ac.il/
> >
> >  Celejar
> >
> > Indeed:  A "Universe Site" that I host on Linode doesn't use https.  So,
> > https://eyeblinkuniverse.com   --   doesn't work:  "Problem loading
> page - Unable to connect"
> > http://eyeblinkuniverse.com--  Works properly on my version of
> Firefox:  Firefox 78.9.0esr (64-bit) on Debian Bullseye.
> >
> > Does this mean that I can look forward to this failure, when the new
> Firefox comes out?  Why can't people still make simple, text-based web
> > pages that can be read to a Blind Person?  See, I don't do forms, data
> entry, or anything on my opening screens.  For that, it links to a
> > Blogspot, which uses some of Google's Bells and Whistles.
> >
> > In other words, is this the end of the "Simple Web"?
>
> https has nothing to do with "simple web" -- I term I don't remember
> having seen before, but I like.  All the TLS negotiation is handled by
> the web server and browser, and it's easy enough to set up with
> letsencryupt that there's really no excuse for not doing it (says the
> guy who hasn't gotten it set up yet for a shotgun club web site he
> manages...  really need to do that...).
>

I use lighttpd for eyeblinkuniverse.com, with nano as my editor. I don't
quite understand the Certificates required for https. I guess it is time
for some lessons.

Thanks!

Kenneth Parker

>

Re: Firefox HTTPS-only mode breaks sites that return 404 for HTTPS connections

[Joe Pfeiffer]

Kenneth Parker  writes:

> On Wed, Apr 14, 2021 at 10:16 PM Celejar  wrote:
>
>  On Wed, 14 Apr 2021 22:57:01 +0100
>  piorunz  wrote:
>
>  > On 14/04/2021 17:19, Celejar wrote:
>  > > Hi,
>  > >
>  > > I recently switched to Firefox's native HTTPS-Only mode from the
>  > > HTTPS Everywhere extension, and I've just made the nasty discovery that
>  > > a bunch of links that had been returning 404, which I had been assuming
>  > > were dead links, were actually perfectly valid pages, which Firefox had
>  > > been "upgrading" to HTTPS, and then getting 404s from web servers that
>  > > weren't offering them via HTTPS (but still apparently accepting
>  > > connections via HTTPS). Clicking on the little lock icon and turning
>  > > HTTPS-Only mode off for the website doesn't seem to have any effect -
>  > > the only thing that lets me actually access these pages is turning off
>  > > HTTPS-Only mode via the general Settings page (or about:config).
>  > >
>  > > When the website doesn't offer HTTPS at all, then Firefox offers to
>  > > connect via HTTP, after a warning, and that's fine. But having pages
>  > > become completely inaccessible is intolerable - I now have to check
>  > > every 404 I get by turning off HTTPS-Only mode and seeing if the
>  > > page is actually there. Am I missing something here, or is HTTPS-Only
>  > > mode just badly broken?
>  > 
>  > It certainly works fine for me. I use https only mode for many months
>  > now. Can you bring an example of a page which returns good page on http,
>  > but 404 error on https?
>
>  http://www.daat.ac.il/
>  https://www.daat.ac.il/
>
>  Celejar
>
> Indeed:  A "Universe Site" that I host on Linode doesn't use https.  So, 
> https://eyeblinkuniverse.com   --   doesn't work:  "Problem loading page - 
> Unable to connect"
> http://eyeblinkuniverse.com--  Works properly on my version of Firefox:  
> Firefox 78.9.0esr (64-bit) on Debian Bullseye.
>
> Does this mean that I can look forward to this failure, when the new Firefox 
> comes out?  Why can't people still make simple, text-based web
> pages that can be read to a Blind Person?  See, I don't do forms, data entry, 
> or anything on my opening screens.  For that, it links to a
> Blogspot, which uses some of Google's Bells and Whistles.
>
> In other words, is this the end of the "Simple Web"?

https has nothing to do with "simple web" -- I term I don't remember
having seen before, but I like.  All the TLS negotiation is handled by
the web server and browser, and it's easy enough to set up with
letsencryupt that there's really no excuse for not doing it (says the
guy who hasn't gotten it set up yet for a shotgun club web site he
manages...  really need to do that...).

Re: Firefox HTTPS-only mode breaks sites that return 404 for HTTPS connections

[Kenneth Parker]

On Wed, Apr 14, 2021 at 10:16 PM Celejar  wrote:

> On Wed, 14 Apr 2021 22:57:01 +0100
> piorunz  wrote:
>
> > On 14/04/2021 17:19, Celejar wrote:
> > > Hi,
> > >
> > > I recently switched to Firefox's native HTTPS-Only mode from the
> > > HTTPS Everywhere extension, and I've just made the nasty discovery that
> > > a bunch of links that had been returning 404, which I had been assuming
> > > were dead links, were actually perfectly valid pages, which Firefox had
> > > been "upgrading" to HTTPS, and then getting 404s from web servers that
> > > weren't offering them via HTTPS (but still apparently accepting
> > > connections via HTTPS). Clicking on the little lock icon and turning
> > > HTTPS-Only mode off for the website doesn't seem to have any effect -
> > > the only thing that lets me actually access these pages is turning off
> > > HTTPS-Only mode via the general Settings page (or about:config).
> > >
> > > When the website doesn't offer HTTPS at all, then Firefox offers to
> > > connect via HTTP, after a warning, and that's fine. But having pages
> > > become completely inaccessible is intolerable - I now have to check
> > > every 404 I get by turning off HTTPS-Only mode and seeing if the
> > > page is actually there. Am I missing something here, or is HTTPS-Only
> > > mode just badly broken?
> >
> > It certainly works fine for me. I use https only mode for many months
> > now. Can you bring an example of a page which returns good page on http,
> > but 404 error on https?
>
> http://www.daat.ac.il/
> https://www.daat.ac.il/
>
> Celejar
>

Indeed:  A "Universe Site" that I host on Linode doesn't use https.  So,
https://eyeblinkuniverse.com   --   doesn't work:  "Problem loading page -
Unable to connect"
http://eyeblinkuniverse.com--  Works properly on my version of
Firefox:  Firefox 78.9.0esr (64-bit) on Debian Bullseye.

Does this mean that I can look forward to this failure, when the new
Firefox comes out?  Why can't people still make simple, text-based web
pages that can be read to a Blind Person?  See, I don't do forms, data
entry, or anything on my opening screens.  For that, it links to a
Blogspot, which uses some of Google's Bells and Whistles.

In other words, is this the end of the "Simple Web"?

Thanks!

Kenneth Parker

Re: Firefox HTTPS-only mode breaks sites that return 404 for HTTPS connections

[Celejar]

On Wed, 14 Apr 2021 16:01:13 -0300
riveravaldez  wrote:

> On 4/14/21, Celejar  wrote:
> > Hi,
> >
> > I recently switched to Firefox's native HTTPS-Only mode from the
> > HTTPS Everywhere extension, and I've just made the nasty discovery that
> > a bunch of links that had been returning 404, which I had been assuming
> > were dead links, were actually perfectly valid pages, which Firefox had
> > been "upgrading" to HTTPS, and then getting 404s from web servers that
> > weren't offering them via HTTPS (but still apparently accepting
> > connections via HTTPS). Clicking on the little lock icon and turning
> > HTTPS-Only mode off for the website doesn't seem to have any effect -
> > the only thing that lets me actually access these pages is turning off
> > HTTPS-Only mode via the general Settings page (or about:config).
> >
> > When the website doesn't offer HTTPS at all, then Firefox offers to
> > connect via HTTP, after a warning, and that's fine. But having pages
> > become completely inaccessible is intolerable - I now have to check
> > every 404 I get by turning off HTTPS-Only mode and seeing if the
> > page is actually there. Am I missing something here, or is HTTPS-Only
> > mode just badly broken?
> >
> > https://blog.mozilla.org/security/2020/11/17/firefox-83-introduces-https-only-mode/
> > https://news.ycombinator.com/item?id=25121604
> >
> > Celejar
> 
> Not sure if useful at all, but I've been using firefox-release (just running
> the binary from Mozilla) as main browser for years, and enabled HTTPS-
> Only Mode upon HTTPS Everywhere add-on/extension simply because
> I had it installed and I was letting it be (?). And that (unexpected?)
> combination seems to work fine (in comparison with your case). Maybe
> you want to try it just to avoid that extra-work until a proper
> solution appears...

Thanks. Perhaps I'll try that.

Celejar

Re: Firefox HTTPS-only mode breaks sites that return 404 for HTTPS connections

[Celejar]

On Wed, 14 Apr 2021 22:57:01 +0100
piorunz  wrote:

> On 14/04/2021 17:19, Celejar wrote:
> > Hi,
> >
> > I recently switched to Firefox's native HTTPS-Only mode from the
> > HTTPS Everywhere extension, and I've just made the nasty discovery that
> > a bunch of links that had been returning 404, which I had been assuming
> > were dead links, were actually perfectly valid pages, which Firefox had
> > been "upgrading" to HTTPS, and then getting 404s from web servers that
> > weren't offering them via HTTPS (but still apparently accepting
> > connections via HTTPS). Clicking on the little lock icon and turning
> > HTTPS-Only mode off for the website doesn't seem to have any effect -
> > the only thing that lets me actually access these pages is turning off
> > HTTPS-Only mode via the general Settings page (or about:config).
> >
> > When the website doesn't offer HTTPS at all, then Firefox offers to
> > connect via HTTP, after a warning, and that's fine. But having pages
> > become completely inaccessible is intolerable - I now have to check
> > every 404 I get by turning off HTTPS-Only mode and seeing if the
> > page is actually there. Am I missing something here, or is HTTPS-Only
> > mode just badly broken?
> 
> It certainly works fine for me. I use https only mode for many months
> now. Can you bring an example of a page which returns good page on http,
> but 404 error on https?

http://www.daat.ac.il/
https://www.daat.ac.il/

Celejar

Re: hello. Please help this big report go in the correct direction....

[Robbi Nespu]

Hello Tim

On 4/14/21 5:04 PM, Timothy Danielson wrote:


strangely I had a similar readout after I attempted to run reportbug but i 
am honeslty a bit foggy on what I did:

timdanielson@td546:~$ google-drive-ocamlfuse ~/mount/google-drive
Could not find the database of available applications, run 
update-command-not-found as root to fix this
Sorry, command-not-found has crashed! Please file a bug report at:
http://www.debian.org/Bugs/Reporting  
Please include the following information with the report:

command-not-found version: 0.3
Python version: 3.7.3 final 0
Distributor ID: PureOS
Description:PureOS
Release:9.0
Codename:   amber
Exception information:

local variable 'cnf' referenced before assignment
Traceback (most recent call last):
   File "/usr/share/command-not-found/CommandNotFound/util.py", line 23, in 
crash_guard
 callback()
   File "/usr/lib/command-not-found", line 93, in main
 if not cnf.advise(args[0], options.ignore_installed) and not 
options.no_failure_msg:
UnboundLocalError: local variable 'cnf' referenced before assignment


I also foggy what when wrong here. But you may issue the problem on 
upstream here https://github.com/astrada/google-drive-ocamlfuse , it 
also have wiki for some reference there.


As I see there is 2 useful command, which are

$ google-drive-ocamlfuse -cc # to clear the cache
$ google-drive-ocamlfuse -debug mountpoint #  turn on debug logging

and 1 directory located on ~/.gdfuse/default may contain something 
interesting for bug reporting


Anyway, it not much of Debian related, try find some help with upstream.

I don't think you can find some help from you distro, because it already 
dead long time ago, check here 
https://lists.debian.org/debian-derivatives/2014/04/msg9.html and 
https://wiki.debian.org/Derivatives/Census/PureOS and 
https://forums.puri.sm/t/when-was-pureos-initially-released/11576


Why not take a backup and switch to Debian? :)

Since you using KDE, you may check 
https://tracker.debian.org/pkg/kio-gdrive


Check and try to understand https://wiki.debian.org/DontBreakDebian 
because I understand most people using Debian derivatives tend to use 
PPA and actually it not encourage and may break your Debian



Likely, upstream google-drive-ocamlfuse seem have clash on licensing too.

If you going to switch to Debian, you may use the unofficial image 
http://cdimage.debian.org/cdimage/unofficial/non-free/images-including-firmware/ 
which already include non-free firmware.



--
Email : Robbi Nespu 
PGP fingerprint : D311 B5FF EEE6 0BE8 9C91 FA9E 0C81 FA30 3B3A 80BA
PGP key : https://keybase.io/robbinespu/pgp_keys.asc

Re: Firefox HTTPS-only mode breaks sites that return 404 for HTTPS connections

[Nito]

On Wed, Apr 14, 2021 at 16:01:13 -0300, riveravaldez wrote:
>
> Not sure if useful at all, but I've been using firefox-release (just running
> the binary from Mozilla) as main browser for years, and enabled HTTPS-
> Only Mode upon HTTPS Everywhere add-on/extension simply because
> I had it installed and I was letting it be (?). And that (unexpected?)
> combination seems to work fine 

Not OP, but a few months ago I did experience a similar problem when setting up 
a reverse-proxy for one of my private domains. For test purposes I initially 
did not set up nginx to deal with TLS for this domain (but others already had 
TLS enabled). Firefox (both with and without HTTPS Everywhere) would always 
switch to TLS, resulting in a cert-mismatch warning and me ending up on a 
different site than intended.
I wasted a few hours trying to figure out what was wrong with my proxy config, 
before I realised (and verified with curl iirc) the fault was with the 
browsers (iinm Chromium messed up similarly).
Once HTTPS was enabled for all domains on this server everything was working 
as expected in browsers again.

--

Nito

Re: pci 0000:00:01:0: MSI quirk detected; subordinated MSI disabled ...

[Albretch Mueller]

> You could then try to boot with the pci=nomsi boot parameter ...

 if I boot up passing to the kernel the start up option:

 knoppix64 debug pci=nomsi noapic

 I would get just two lines further bellow:

 pci :00:01:0: MSI quirk detected; subordinated MSI disabled
 PCI Interrupt Link [LNKA] enabled at IRQ 11
 PCI: setting IRQ 11 as level-triggered

 and the boot process would stop.

> there is a chance that this message is an information message about a step of 
> the boot procédure that just has been successful and that the problematic 
> step that freezes your computer is the next one...

 is there a way for me to know what the next step would be? The kernel
boot up logs should be " (do)ing ", "... ok"

 I know the boot process is different for each motherboard, a bit of
googling gave me:

 http://forums.debian.net/viewtopic.php?f=5=47164

[ ...] ACPI: PCI Interrupt Link [LNKA] enabled at IRQ 11
[ ...] ACPI: PCI Interrupt :00:11.1[A] -> Link [LNKA] -> GSI 11
(level, low) -> IRQ 11
[ ...] VIA_IDE :00:11.1: VIA VLink IRQ fixup, from 255 to 11
[ ...] VP_IDE: not 100% native mode: will probe irqs later

 so the problem might be related to the video card?

 I also found this message at an ubuntu forum:

 https://ubuntuforums.org/showthread.php?t=2202279

 Re: New PC, cannot install Ubuntu (pci :00:01.0: MSI quirk detected)

Oldfred, I want to thank you for your tireless effort to help me
with resolving this issue. I am very grateful that people like you are
available to the community to support noobs like myself

I am posting this from my new Xubuntu install, so it seems we
finally managed to get things working!!!

Here is a recap from the steps I took that (I think) lead to the resolution:
- (Obviously) go through the posts provided by Oldfred, tried out
several options
- I found some info pointing towards USB legacy support and MSI quirk
- Disabled legacy support in my BIOS
- Got that Checking NVRAM error
- Reset my CMOS (through the jumper setting described in my
motherboard manual)
- Disabled legacy support again, checked some other BIOS settings
to make sure they are what they were before resetting the CMOS
- Set my CD/DVD up as primary boot device (in BIOS)
- Used a (X)Ubuntu installation CD instead of a USB (as legacy
support is now disabled)
- Up till this point I still had that Checking NVRAM error at POST.
- Almost went out to buy a new motherboard, gave it one more try
by unplugging my USB keyboard and mouse, booting with Xubuntu
installation CD (12.04.3 LTS)
- BINGO! Installation starts!!!
- During installation, plug back in my USB keyboard and mouse
- Let it complete
- Restart is needed in order to complete installation
- Had some more issues after that with keyboard not working in GRUB menu
- Somehow magically ended up in my BIOS again (by using the DEL
key during boot)
- Enabled USB legacy support
- Reboot
- Finally able to have Xubuntu boot again, login, install latest
updates, working instllation!

I am hoping that I described all this in enough detail so that it
can help others with similar issues.

In the end it seems that when you get that MSI quirk error,
disabling USB legacy support in your BIOS goes a long way. That
Checking NVRAM error is (or looks to be) because you still have USB
devices plugged in when USB legacy support is disabled.

So resetting the CMOS may not have been necessary after all. I did
not have a PS/2 keyboard/mouse available, but probably disabling
legacy USB support in the BIOS and using PS/2 keyboard+mouse would
have worked directly.
~
 but I very much doubt that MSI could be in business requiring so much
work for regular folks out there.

 I will keep a log here about how I solved that problem in order to
help people running into the same hurdle

 lbrtchx

Re: Firefox HTTPS-only mode breaks sites that return 404 for HTTPS connections

[piorunz]

On 14/04/2021 17:19, Celejar wrote:

Hi,

I recently switched to Firefox's native HTTPS-Only mode from the
HTTPS Everywhere extension, and I've just made the nasty discovery that
a bunch of links that had been returning 404, which I had been assuming
were dead links, were actually perfectly valid pages, which Firefox had
been "upgrading" to HTTPS, and then getting 404s from web servers that
weren't offering them via HTTPS (but still apparently accepting
connections via HTTPS). Clicking on the little lock icon and turning
HTTPS-Only mode off for the website doesn't seem to have any effect -
the only thing that lets me actually access these pages is turning off
HTTPS-Only mode via the general Settings page (or about:config).

When the website doesn't offer HTTPS at all, then Firefox offers to
connect via HTTP, after a warning, and that's fine. But having pages
become completely inaccessible is intolerable - I now have to check
every 404 I get by turning off HTTPS-Only mode and seeing if the
page is actually there. Am I missing something here, or is HTTPS-Only
mode just badly broken?


It certainly works fine for me. I use https only mode for many months
now. Can you bring an example of a page which returns good page on http,
but 404 error on https?

--

With kindest regards, piorunz.

⢀⣴⠾⠻⢶⣦⠀
⣾⠁⢠⠒⠀⣿⡁ Debian - The universal operating system
⢿⡄⠘⠷⠚⠋⠀ https://www.debian.org
⠈⠳⣄

Re: Montar HD no boot automaticamente com permissão para usuários comuns

[Humberto A. Sousa]

De boa!

Só seria legal se o Juscelino tivesse respondido para a lista ao invés 
de responder no particular.




Saudações,



Humberto Araujo de Sousa
humbe...@dontec.com.br
ICQ: @hasousa
https://mars.nasa.gov/layout/embed/send-your-name/mars2020/certificate/?cn=57592057281

Em 14/04/2021 13:46, China escreveu:

A resposta do Juscelino foi a solução para mim. Eu já tinha criado o
ponto de montagem, mas não funcionou declarando o inode do HD como
antigamente. Coloquei o UID e tá de boa. Obrigado!

Em ter., 13 de abr. de 2021 às 15:17, Juscelino Cordeiro
 escreveu:


Tenta executar o comando sudo blkid e pegar o uid do disco. Coloca no fstab no 
lugar de /dev/sda5.
Talvez seja necessário criar a pasta onde vc quer o hd seja montado.

Em ter, 13 de abr de 2021 14:48, China  escreveu:


Boa tarde.

Instalei um SSD no notebook e o HD mecanico continua na máquina para
ser usado como depósito de arquivos. O método que aparece para o
acesso é clicar no Nautilus em "Outros locais" e clicar no HD, daí
insere a senha de usuário e o HD é montado na pasta
/media/$user/UID_do_HD. Não que este método não me atenda, mas quis
fazer do jeito que conheço que é incluir a linha de montagem no
/etc/fstab:
  "/dev/sda5 /home/china/Arquivos ext4 defauts 0 0"

Com esta linha esperava o HD montado na pasta "Arquivos" dentro do meu
/home, mas não funcionou, a máquina nem deu boot alegando violação do
journal e parou na tela de manutenção. Só consegui dar o boot após
comentar esta linha.

No Google tem vários tutoriais ensinando este método e outros como
usar a ferramenta "Discos" do Gnome, mas não funcionou para mim. Então
pergunto: como montar o HD no boot que fique dentro do meu /home com
permissões de usuário sem ter de clicar em lugar nenhum e colocar
senha para montar?

Ambiente Debian stable com kernel 5.10 do stech-backports.

--
Enviado de um dispositivo móvel

Re: [OT] Help with clonezilla-live: LUKS/LVM2 image backup without partclone.dd

[l0f4r0]

Hi Didier,

Thanks for your answer.

14 avr. 2021, 20:50 de didier.gau...@gmail.com:

> So it seems to me appropriate that dd is used as you want an image of a whole 
> disk?
>
Whole disk here doesn't mean 1 image file.

Actually, after running clonezilla I get numerous files:

nvme0n1p1.vfat-ptcl-img.zst.aa
nvme0n1p2.dd-ptcl-img.zst.aa
nvme0n1p3.ext4-ptcl-img.zst.aa
nvme0n1p4.ntfs-ptcl-img.zst.aa
nvme0n1p5.dd-ptcl-img.zst.aa
nvme0n1p5.dd-ptcl-img.zst.ab
nvme0n1p5.dd-ptcl-img.zst.ac
[...]

Partitions are processed one by one by clonezilla.
That's why I said that partitions like nvme0n1p{1,3,4} are backupped with 
`partclone.{vfat,ext4,ntfs}` respectively, while nvme0n1p{2,5} are backupped 
with `partclone.dd`.

Logs show that each step has run successfully:

Finished saving /dev/nvme0n1p1 as 
/tmp/ecryptfs_mnt.VDFrH7/nvme0n1p1.vfat-ptcl-img.zst
Finished saving /dev/nvme0n1p2 as 
/tmp/ecryptfs_mnt.VDFrH7/nvme0n1p2.dd-ptcl-img.zst
Finished saving /dev/nvme0n1p3 as 
/tmp/ecryptfs_mnt.VDFrH7/nvme0n1p3.ext4-ptcl-img.zst
Finished saving /dev/nvme0n1p4 as 
/tmp/ecryptfs_mnt.VDFrH7/nvme0n1p4.ntfs-ptcl-img.zst
Finished saving /dev/nvme0n1p5 as 
/tmp/ecryptfs_mnt.VDFrH7/nvme0n1p5.dd-ptcl-img.zst

Best regards,
l0f4r0

Re: Mysql et php5.6

[Jean-Michel OLTRA]

Bonjour,


Le mercredi 14 avril 2021, ajh-valmer a écrit...


> > Erreur, ou pas ? Avec connect_errno et connect_error, tu devrais en savoir
> > un peu plus sur l'erreur de connexion :

> Ou et comment placer ces scripts ?
> connect_errno , connect_error

Tiens, une doc en français :
https://www.php.net/manual/fr/mysqli.quickstart.connections.php

> > Sinon, tu as également les logs de mysql/mariadb, si tu y as accès.

> C'est une base de données MySQL.

Ben, c'est un peu pareil. Si il y a un souci de connexion, l'erreur est
peut-être mentionnée dedans. Il te faudrait lire ces logs.

> > A voir si ce ne serait pas un souci de connexion sur le socket unix sur
> > localhost vs la connexion tcp sur le port 3306 ?

> Tous les fichiers de mysql indiquent bien le port 3306.

Tu devrais effectuer la requête suivante sur la base mysql :

MariaDB [mysql]> select user, plugin, host from user;
(chez moi, c'est MariaDB, pas mysql).

J'obtiens ça :

+-+---+---+
| User| plugin| Host  |
+-+---+---+
| root| unix_socket   | localhost |
| jm  | mysql_native_password | localhost |
| mariadb.sys | mysql_native_password | localhost |
+-+---+---+

Et lire ce qui concerne l'authentification sous mysql 8.
Ton souci vient peut-être (probablement) de là.

-- 
jm

Re: GRUB no inicia sistema antiguo (era: debian 11 kde primer arranque)

[Camaleón]

El 2021-04-14 a las 15:20 -0300, Marcelo Giordano escribió:

> Ah... vale, ya sé lo que pasa, al descargarlo se elimina
> > automáticamente. No me di cuenta de su funcionamiento, qué cosas hay por
> > la web :-)
> > 
> > Prueba con un servicio clásico como https://wetransfer.com/, y dile que
> > te genere un enlace para enviarlo a la lista.
> > 
> Ahi va.
> 
> Gracias
> 
> https://wetransfer.com/downloads/cd700a6e3c7819376cc4252355d78adf20210414181922/4bd7e6d49468e108e7b753e65959c0dc20210414181941/19fde6
>  
> 

Perfecto, ahora lo veo completo, gracias.

Bueno, lo primero que me llama la atención es que las dos entradas para 
Debian 11 y Debian 10 apuntan a la misma partición raíz (mismo UUID), 
lo cual no es correcto:

Debian 11:
linux   /vmlinuz-5.10.0-5-amd64 root=UUID=c69220c8-b56c-482a-ba9f-8189ca315311 
ro  quiet

Debian 10:
linux   /vmlinuz-4.19.0-16-amd64 root=UUID=c69220c8-b56c-482a-ba9f-8189ca315311 
ro  quiet
 
Inicia Debian 11, ejecuta «cat /etc/fstab» y como súperusuario, «blkid» 
y copia/pega en el mismo correo la salida de las dos órdenes, que no 
serán muy largas, para ver cómo tiene el sistema identificadas las particiones.

Saludos,

-- 
Camaleón 

Re: Not Sure Which Package to Submit Bug Against

[deloptes]

David Wright wrote:

> On Tue 13 Apr 2021 at 00:49:48 (+0200), deloptes wrote:
>> Kent West wrote:
>> 

>> 
>> Perhaps you try locking the session of the first user and see what
>> happens - if the second has access to the audio.
> 
> It would also be interesting to know what happens when the first
> person logs in remotely and a second person then logs in at the
> console. Does the first person get ownership at all, and if so,
> does the second person wrest it from them?
> 

I read the documentation around ConsoleKit and systemd-login (which is more
or less the same) briefly
A difference is made by console login and remote login. All this
restrictions are related to local login, where the user occupies the
hardware, because obviously only one user can sit infront of the PC.
What I wanted to try, but did not have the time to do is to test all of this
after installing ConsoleKit ... or better wait for Debian 11.

Re: Firefox HTTPS-only mode breaks sites that return 404 for HTTPS connections

[riveravaldez]

On 4/14/21, Celejar  wrote:
> Hi,
>
> I recently switched to Firefox's native HTTPS-Only mode from the
> HTTPS Everywhere extension, and I've just made the nasty discovery that
> a bunch of links that had been returning 404, which I had been assuming
> were dead links, were actually perfectly valid pages, which Firefox had
> been "upgrading" to HTTPS, and then getting 404s from web servers that
> weren't offering them via HTTPS (but still apparently accepting
> connections via HTTPS). Clicking on the little lock icon and turning
> HTTPS-Only mode off for the website doesn't seem to have any effect -
> the only thing that lets me actually access these pages is turning off
> HTTPS-Only mode via the general Settings page (or about:config).
>
> When the website doesn't offer HTTPS at all, then Firefox offers to
> connect via HTTP, after a warning, and that's fine. But having pages
> become completely inaccessible is intolerable - I now have to check
> every 404 I get by turning off HTTPS-Only mode and seeing if the
> page is actually there. Am I missing something here, or is HTTPS-Only
> mode just badly broken?
>
> https://blog.mozilla.org/security/2020/11/17/firefox-83-introduces-https-only-mode/
> https://news.ycombinator.com/item?id=25121604
>
> Celejar

Not sure if useful at all, but I've been using firefox-release (just running
the binary from Mozilla) as main browser for years, and enabled HTTPS-
Only Mode upon HTTPS Everywhere add-on/extension simply because
I had it installed and I was letting it be (?). And that (unexpected?)
combination seems to work fine (in comparison with your case). Maybe
you want to try it just to avoid that extra-work until a proper
solution appears...

Best regards.

Re: [OT] Help with clonezilla-live: LUKS/LVM2 image backup without partclone.dd

[didier gaumet]

Hello,

Disclaimer: I have never used Clonezilla nor LUKS

"Many File systems are supported: (1) ext2, ext3, ext4, reiserfs, 
reiser4, xfs, jfs, btrfs, f2fs and nilfs2 of GNU/Linux, (2) FAT12, 
FAT16, FAT32, NTFS of MS Windows, (3) HFS+ of Mac OS, (4) UFS of 
FreeBSD, NetBSD, and OpenBSD, (5) minix of Minix, and (6) VMFS3 and 
VMFS5 of VMWare ESX. Therefore you can clone GNU/Linux, MS windows, 
Intel-based Mac OS, FreeBSD, NetBSD, OpenBSD, Minix, VMWare ESX and 
Chrome OS/Chromium OS, no matter it's 32-bit (x86) or 64-bit (x86-64) 
OS. For these file systems, only used blocks in partition are saved and 
restored by Partclone. For unsupported file system, sector-to-sector 
copy is done by dd in Clonezilla."


So it seems to me appropriate that dd is used as you want an image of a 
whole disk?

Re: pci 0000:00:01:0: MSI quirk detected; subordinated MSI disabled ...

[didier gaumet]

Hello,

I think that this error message is not about MSI-the-manufacturer but 
that here MSI stands for Message Signaled Interrupt:

 https://www.kernel.org/doc/html/latest/PCI/msi-howto.html

You could then try to boot with the pci=nomsi boot parameter to see if 
your PC boots properly. But there is a chance that this message is an 
information message about a step of the boot procédure that just has 
been succesful and that the problematic step that freezes your computer 
is the next one...

Re: GRUB no inicia sistema antiguo (era: debian 11 kde primer arranque)

[Marcelo Giordano]

Ah... vale, ya sé lo que pasa, al descargarlo se elimina

automáticamente. No me di cuenta de su funcionamiento, qué cosas hay por
la web :-)

Prueba con un servicio clásico como https://wetransfer.com/, y dile que
te genere un enlace para enviarlo a la lista.

Saludos,


Ahi va.

Gracias

https://wetransfer.com/downloads/cd700a6e3c7819376cc4252355d78adf20210414181922/4bd7e6d49468e108e7b753e65959c0dc20210414181941/19fde6 

[OT] Help with clonezilla-live: LUKS/LVM2 image backup without partclone.dd

[l0f4r0]

Hi,

Do some of you use clonezilla for your backups, especially with a LVM2/LUKS 
configuration please?

I've written some days ago to the clonezilla-live mailing-list but I'm not sure 
I'll ever get answers considering the very very low traffic on it. So maybe 
I'll be more lucky here between Debianists :)

Here is the partition configuration on my SSD:

NAME   FSTYPE  LABEL MOUNTPOINT 
SIZE
nvme0n1 
477G
├─nvme0n1p1    vfat  /boot/efi  
477M
├─nvme0n1p2 
 16M
├─nvme0n1p3    ext4    boot  /boot  
477M
├─nvme0n1p4    ntfs    WinRE_DRV   
1000M
└─nvme0n1p5    crypto_LUKS    
474,8G
  └─nvme0n1p5_crypt    LVM2_member    
474,8G
    ├─myLvmGroup-myLvmSwap swap  [SWAP]    
14,9G
    ├─myLvmGroup-myLvmVar  ext4    var   /var  
18,6G
    ├─myLvmGroup-myLvmRoot ext4    root  / 
74,5G
    └─myLvmGroup-myLvmHome ext4    home  /home  
326G

I want to create a clonezilla image of everything on an external drive 
partition.

So I boot on a clonezilla-live installed on a USB stick, open a shell in order 
to open my LUKS container (at this stage myLvmGroup is recognized and 
deployed), start clonezilla and use the ncurses interface.

Partitions like nvme0n1p{1,3,4} are backupped cleverly with 
`partclone.{vfat,ext4,ntfs}` respectively. Good.

`partclone.dd` is used for nvme0n1p2 (no big deal, it's an Ms reserved 
partition). OK.

However, `partclone.dd` is used as well with nvme0n1p5. Not good. This is not 
very handy (it's time-consuming and produces a large file as `dd` doesn't 
understand the underlying fs with the empty space).

I think this is due to the fact that my LVM is shutdowned in clonezilla before 
the backup starts, as indicated in my logs:

Clonezilla image dir: /home/partimag
Shutting down the Logical Volume Manager
  Shutting Down logical volume: /dev/myLvmGroup/myLvmHome
  Shutting Down logical volume: /dev/myLvmGroup/myLvmRoot
  Shutting Down logical volume: /dev/myLvmGroup/myLvmSwap
  Shutting Down logical volume: /dev/myLvmGroup/myLvmVar
  Shutting Down volume group: myLvmGroup
Finished Shutting down the Logical Volume Manager

* Do you know why there is this behavior please?

* Am I supposed to stay in shell mode and use `partclone` directly on each of 
my logical volumes? Like the following maybe:
`partclone.ext4 -z 10485760 -N -L /home/partimg/partclone.log -c -s 
/dev/mapper/myLvmGroup-myLvmRoot --output - | zstdmt -c -3 --rsyncable | split 
-a 2 -b 4096MB - /home/partimg/myLvmGroup-myLvmRoot.ext4-ptcl-img.zst. 2> 
/home/partimg/split_error.log`?

* I haven't done it so far but do you think I can use the following command 
directly
`/usr/sbin/ocs-sr -q2 -c -j2 -z9p -i 4096 -fsck -enc -p choose savedisk 
fulldisk_encrypted_img nvme0n1`? Or will my LVM be shutdowned as well?

* Do you have any advice to share?

Thank you in advance :)

Best regards,
l0f4r0

Participe dos canais de notícias sobre Debian em português

[Paulo Henrique de Lima Santana]

Olá pessoal.

Estamos reativando dois canais para envio de notícias em português sobre 
o Debian.


Canal no telegram "Debian Brasil Notícias":
https://t.me/DebianBrasilNoticias

Lista de discusão "Debian News Portuguese"
https://lists.debian.org/debian-news-portuguese

Se tivermos muitos inscritos nesses dois canais, poderemos passar a 
enviar as notícias em um canal e uma lista ao invés de enviar para 
várioss grupos e listas como temos que fazer hoje :-)


Entrem lá!

Abraços,

--
Paulo Henrique de Lima Santana (phls)
Curitiba - Brasil
Debian Developer
Diretor do Instituto para Conservação de Tecnologias Livres
Site: http://phls.com.br
GNU/Linux user: 228719  GPG ID: 0443C450



OpenPGP_signature
Description: OpenPGP digital signature

Re: Not Sure Which Package to Submit Bug Against

[Dan Ritter]

Nicholas Geovanis wrote: 
> 
> The most interesting question to me:
> Has doc or first-principles failed so badly that we resort to
> experimentation again?

Yes, in the sense that there are too many ways for this to
legally exist to be able to say: "package foo is causing this,
and you can either change the config this way, install
foo-alternative, or use this different hardware".

-dsr-

Re: Not Sure Which Package to Submit Bug Against

[Nicholas Geovanis]

On Wed, Apr 14, 2021, 11:20 AM David Wright 
wrote:

> On Tue 13 Apr 2021 at 00:49:48 (+0200), deloptes wrote:
> > Kent West wrote:
> >
> > > I did some experimentation afterwards, and have discovered that if
>


> > Perhaps you try locking the session of the first user and see what
> happens -
> > if the second has access to the audio.
>
> It would also be interesting to know what happens when the first
> person logs in remotely and a second person then logs in at the
> console. Does the first person get ownership at all, and if so,
> does the second person wrest it from them?
>

The most interesting question to me:
Has doc or first-principles failed so badly that we resort to
experimentation again?
:-)


Cheers,
> David.
>
>

Re: Montar HD no boot automaticamente com permissão para usuários comuns

[China]

A resposta do Juscelino foi a solução para mim. Eu já tinha criado o
ponto de montagem, mas não funcionou declarando o inode do HD como
antigamente. Coloquei o UID e tá de boa. Obrigado!

Em ter., 13 de abr. de 2021 às 15:17, Juscelino Cordeiro
 escreveu:
>
> Tenta executar o comando sudo blkid e pegar o uid do disco. Coloca no fstab 
> no lugar de /dev/sda5.
> Talvez seja necessário criar a pasta onde vc quer o hd seja montado.
>
> Em ter, 13 de abr de 2021 14:48, China  escreveu:
>>
>> Boa tarde.
>>
>> Instalei um SSD no notebook e o HD mecanico continua na máquina para
>> ser usado como depósito de arquivos. O método que aparece para o
>> acesso é clicar no Nautilus em "Outros locais" e clicar no HD, daí
>> insere a senha de usuário e o HD é montado na pasta
>> /media/$user/UID_do_HD. Não que este método não me atenda, mas quis
>> fazer do jeito que conheço que é incluir a linha de montagem no
>> /etc/fstab:
>>  "/dev/sda5 /home/china/Arquivos ext4 defauts 0 0"
>>
>> Com esta linha esperava o HD montado na pasta "Arquivos" dentro do meu
>> /home, mas não funcionou, a máquina nem deu boot alegando violação do
>> journal e parou na tela de manutenção. Só consegui dar o boot após
>> comentar esta linha.
>>
>> No Google tem vários tutoriais ensinando este método e outros como
>> usar a ferramenta "Discos" do Gnome, mas não funcionou para mim. Então
>> pergunto: como montar o HD no boot que fique dentro do meu /home com
>> permissões de usuário sem ter de clicar em lugar nenhum e colocar
>> senha para montar?
>>
>> Ambiente Debian stable com kernel 5.10 do stech-backports.
>>
>> --
>> Enviado de um dispositivo móvel
>>


-- 

Enviado de um dispositivo móvel

Re: Re: trouble with sbuild

[Patrice Duroux]

I should have stacked the changelog to reads: wiki, manpages, README!
Now I will think about a wish report to sbuild package.
Thanks

Re: GRUB no inicia sistema antiguo (era: debian 11 kde primer arranque)

[Camaleón]

El 2021-04-14 a las 12:31 -0300, Marcelo Giordano escribió:
> 
> On 14/4/21 12:28, Camaleón wrote:
> > El 2021-04-14 a las 11:09 -0300, Marcelo Giordano escribió:
> > > On 14/4/21 02:48, Camaleón wrote:
> > > > > > Inicia Debian 11 normalmente y manda a la lista o sube a algún sitio
> > > > > > tipo pastebin (p. ej., «https://paste.debian.net/»), el contenido 
> > > > > > del
> > > > > > archivo de configuracuón de GRUB2 que tienes  en 
> > > > > > «/boot/grub/grub.cfg».
> > > > > > 
> > > > > > Recuerda que también puedes intentar iniciar tu Debian 10 desde la
> > > > > > línea de órdenes de GRUB2. Es un método un poco engorroso porque 
> > > > > > tendrás
> > > > > > que ir probando opciones pero funciona bien:
> > > > > > 
> > > > > > How to Rescue a Non-booting GRUB 2 on Linux
> > > > > > https://www.linux.com/training-tutorials/how-rescue-non-booting-grub-2-linux/
> > > > > > 
> > > > > > Saludos,
> > > > > > 
> > > > > #
> > > > > # DO NOT EDIT THIS FILE
> > > > > #
> > > > > # It is automatically generated by grub-mkconfig using templates
> > > > > # from /etc/grub.d and settings from /etc/default/grub
> > > > > #
> > > > (...)
> > > > 
> > > > Marcelo, el archivo que has enviado está incompleto (cortado), faltan
> > > > las líneas de arranque de los sistemas (menuentry).
> > > > 
> > > > Puedes enviarlo de nuevo a la lista (asegúrate de que copias/pegas
> > > > todas las líneas) o subir el archivo a algún sitio (como
> > > > https://www.file.io).
> > > > 
> > > Que bobo, pasa que lo copié a mano. Que desastre.
> > > 
> > > Gracias por todo, ahí va el link
> > > 
> > > https://file.io/1buBSF9uwRXV 
> > Al acceder me dice que el archivo se ha eliminado :-)
> > 
> > ***
> > The file you requested has been deleted
> > ***
> > 
> > 
> Que raro
> 
> https://file.io/zt4epniMa9rS 
> 
> este link lo probé

Ah... vale, ya sé lo que pasa, al descargarlo se elimina 
automáticamente. No me di cuenta de su funcionamiento, qué cosas hay por
la web :-)

Prueba con un servicio clásico como https://wetransfer.com/, y dile que 
te genere un enlace para enviarlo a la lista.

Saludos,

-- 
Camaleón 

Re: Not Sure Which Package to Submit Bug Against

[David Wright]

On Tue 13 Apr 2021 at 00:49:48 (+0200), deloptes wrote:
> Kent West wrote:
> 
> > I did some experimentation afterwards, and have discovered that if
> > user X mutes the mic, the mic then seems to be "owned" by user X, and
> > no one and no OS can seem to unmute it. I was using the machine as
> > user Y, which is why I couldn't unmute it, and when I handed the
> > laptop to the hardware tech, logged in as user X, bingo! He umuted it.
> 
> I was looking into the same recently and reached the point of systemd +
> ConsoleKit
> 
> Quote: "ConsoleKit ConsoleKit is a framework for defining and tracking
> users, login sessions, and seats. ConsoleKit is currently not actively
> maintained. The focus has shifted to the built-in seat/user/session
> management of Software/systemd called systemd-logind!"
> 
> https://www.freedesktop.org/wiki/Software/ConsoleKit/
> 
> Basically the first logged in user takes ownership of the hardware and
> ConsoleKit or systemd utility should handle the ownership if a second user
> logged in.
> I have not come so far to test this to the end.
> 
> Perhaps you try locking the session of the first user and see what happens -
> if the second has access to the audio.

It would also be interesting to know what happens when the first
person logs in remotely and a second person then logs in at the
console. Does the first person get ownership at all, and if so,
does the second person wrest it from them?

Cheers,
David.

Firefox HTTPS-only mode breaks sites that return 404 for HTTPS connections

[Celejar]

Hi,

I recently switched to Firefox's native HTTPS-Only mode from the
HTTPS Everywhere extension, and I've just made the nasty discovery that
a bunch of links that had been returning 404, which I had been assuming
were dead links, were actually perfectly valid pages, which Firefox had
been "upgrading" to HTTPS, and then getting 404s from web servers that
weren't offering them via HTTPS (but still apparently accepting
connections via HTTPS). Clicking on the little lock icon and turning
HTTPS-Only mode off for the website doesn't seem to have any effect -
the only thing that lets me actually access these pages is turning off
HTTPS-Only mode via the general Settings page (or about:config).

When the website doesn't offer HTTPS at all, then Firefox offers to
connect via HTTP, after a warning, and that's fine. But having pages
become completely inaccessible is intolerable - I now have to check
every 404 I get by turning off HTTPS-Only mode and seeing if the
page is actually there. Am I missing something here, or is HTTPS-Only
mode just badly broken?

https://blog.mozilla.org/security/2020/11/17/firefox-83-introduces-https-only-mode/
https://news.ycombinator.com/item?id=25121604

Celejar

Re: pci 0000:00:01:0: MSI quirk detected; subordinated MSI disabled ...

[David Wright]

On Wed 14 Apr 2021 at 11:26:53 (-0400), Albretch Mueller wrote:
>  is the error message I get when I try to start a PC based on an MSI
> motherboard, with both Debian live an Knoppix (which is a bit more
> flexible about start up options)
> 
>  How can I get passed that error in order to be able to use and
> install Debian live or Knoppix?
> 
> I have tried a number of options such as acpi=off combined with some
> other ones as well as read the MSI-HOWTO.txt
> 
>  
> https://elixir.bootlin.com/linux/v2.6.28/source/Documentation/PCI/MSI-HOWTO.txt
> 
>  on which they talk about "MSI quirks" on chapter 6th, but I can't see
> the relationship with booting Debian Live, so I I haven't been able to
> fix that problem.

You could try disabling what you can in the BIOS to see if you can
boot. Then restore things one by one. That worked for someone—their
problem was caused by USB Legacy Emulation being enabled.

Cheers,
David.

Re: Mysql et php5.6

[ajh-valmer]

On Wednesday 14 April 2021 14:06:07 Jean-Michel OLTRA wrote:
> Erreur, ou pas ? Avec connect_errno et connect_error, tu devrais en savoir
> un peu plus sur l'erreur de connexion :

Ou et comment placer ces scripts ?
connect_errno , connect_error

> Sinon, tu as également les logs de mysql/mariadb, si tu y as accès.

C'est une base de données MySQL.

> A voir si ce ne serait pas un souci de connexion sur le socket unix sur
> localhost vs la connexion tcp sur le port 3306 ?

Tous les fichiers de mysql indiquent bien le port 3306.

Re: GRUB no inicia sistema antiguo (era: debian 11 kde primer arranque)

[Marcelo Giordano]

On 14/4/21 12:28, Camaleón wrote:

El 2021-04-14 a las 11:09 -0300, Marcelo Giordano escribió:

On 14/4/21 02:48, Camaleón wrote:

Inicia Debian 11 normalmente y manda a la lista o sube a algún sitio
tipo pastebin (p. ej., «https://paste.debian.net/»), el contenido del
archivo de configuracuón de GRUB2 que tienes  en «/boot/grub/grub.cfg».

Recuerda que también puedes intentar iniciar tu Debian 10 desde la
línea de órdenes de GRUB2. Es un método un poco engorroso porque tendrás
que ir probando opciones pero funciona bien:

How to Rescue a Non-booting GRUB 2 on Linux
https://www.linux.com/training-tutorials/how-rescue-non-booting-grub-2-linux/

Saludos,


#
# DO NOT EDIT THIS FILE
#
# It is automatically generated by grub-mkconfig using templates
# from /etc/grub.d and settings from /etc/default/grub
#

(...)

Marcelo, el archivo que has enviado está incompleto (cortado), faltan
las líneas de arranque de los sistemas (menuentry).

Puedes enviarlo de nuevo a la lista (asegúrate de que copias/pegas
todas las líneas) o subir el archivo a algún sitio (como
https://www.file.io).


Que bobo, pasa que lo copié a mano. Que desastre.

Gracias por todo, ahí va el link

https://file.io/1buBSF9uwRXV 

Al acceder me dice que el archivo se ha eliminado :-)

***
The file you requested has been deleted
***

Saludos,


Que raro

https://file.io/zt4epniMa9rS 

este link lo probé

Gracias

Re: GRUB no inicia sistema antiguo (era: debian 11 kde primer arranque)

[Camaleón]

El 2021-04-14 a las 11:09 -0300, Marcelo Giordano escribió:
> On 14/4/21 02:48, Camaleón wrote:
> > > > 
> > > > Inicia Debian 11 normalmente y manda a la lista o sube a algún sitio
> > > > tipo pastebin (p. ej., «https://paste.debian.net/»), el contenido del
> > > > archivo de configuracuón de GRUB2 que tienes  en «/boot/grub/grub.cfg».
> > > > 
> > > > Recuerda que también puedes intentar iniciar tu Debian 10 desde la
> > > > línea de órdenes de GRUB2. Es un método un poco engorroso porque tendrás
> > > > que ir probando opciones pero funciona bien:
> > > > 
> > > > How to Rescue a Non-booting GRUB 2 on Linux
> > > > https://www.linux.com/training-tutorials/how-rescue-non-booting-grub-2-linux/
> > > > 
> > > > Saludos,
> > > > 
> > > #
> > > # DO NOT EDIT THIS FILE
> > > #
> > > # It is automatically generated by grub-mkconfig using templates
> > > # from /etc/grub.d and settings from /etc/default/grub
> > > #
> > (...)
> > 
> > Marcelo, el archivo que has enviado está incompleto (cortado), faltan
> > las líneas de arranque de los sistemas (menuentry).
> > 
> > Puedes enviarlo de nuevo a la lista (asegúrate de que copias/pegas
> > todas las líneas) o subir el archivo a algún sitio (como
> > https://www.file.io).
> > 
> Que bobo, pasa que lo copié a mano. Que desastre.
> 
> Gracias por todo, ahí va el link
> 
> https://file.io/1buBSF9uwRXV 

Al acceder me dice que el archivo se ha eliminado :-)

***
The file you requested has been deleted
***

Saludos,

-- 
Camaleón 

pci 0000:00:01:0: MSI quirk detected; subordinated MSI disabled ...

[Albretch Mueller]

 is the error message I get when I try to start a PC based on an MSI
motherboard, with both Debian live an Knoppix (which is a bit more
flexible about start up options)

 How can I get passed that error in order to be able to use and
install Debian live or Knoppix?

I have tried a number of options such as acpi=off combined with some
other ones as well as read the MSI-HOWTO.txt

 https://elixir.bootlin.com/linux/v2.6.28/source/Documentation/PCI/MSI-HOWTO.txt

 on which they talk about "MSI quirks" on chapter 6th, but I can't see
the relationship with booting Debian Live, so I I haven't been able to
fix that problem.

 lbrtchx

Re: SOLUCIONADO: Xerox B215 configurar scanner con TCP/IP en Debian 10 xfce

[Marcelo Giordano]

Y gracias por cerrar el hilo y compartir.

JAP

Por gente como vos funciona esta lista y disfrutamos de usar Debian. Eso 
nos une a todos aunque algunos sean expertos como vos o novatos como yo

Re: Mysql et php5.6

[MERLIN Philippe]

Le mercredi 14 avril 2021, 11:31:49 CEST ajh-valmer a écrit :
> On Tuesday 13 April 2021 00:06:27 Jean-Michel OLTRA wrote:
> > Le lundi 12 avril 2021, ajh-valmer a écrit...
> > 
> > > > > if (!$CNX = mysql_connect($DB_BASE, $DB_USER, $DB_PASS, $DB_SERVER))
> > > > > {
> > > 
> > > La syntaxe ci-dessus n'est pas supprimée en php 5.6.
> > > Normalement, il devrait accepter les scripts php 5
> > 
> > Que mysql_connect existe encore c'est une chose.
> > Que l'ordre des paramètres que tu passes dans ce que tu as écrit soit
> > valide, ça en est une autre. A moins que tu aies mal recopié ton véritable
> > code de connexion ?
> > Ceci étant, je n'utilise pas mysql_connect,
> > mais "new mysqli(server, username, password, db)".
> 
> Ce code ne fonctionne pas non plus.
> Il n'y a pas d'erreur.
> 
> J'ai lu qu'il fallait utiliser mysqli ou PDO à partir de la version 7.
> Mais mysqli me renvoie erreur.
> 
> Bonne journée.

Es tu sur de ta syntaxe, si je regarde la doc sur mysql_connect l'ordre des 
paramètres est : serveur, base .
A part cela je n'utilise pas PHP.
Amitiés.
Philippe Merlin

Re: GRUB no inicia sistema antiguo (era: debian 11 kde primer arranque)

[Marcelo Giordano]

On 14/4/21 02:48, Camaleón wrote:

El 2021-04-13 a las 20:01 -0300, Marcelo Giordano escribió:


On 13/4/21 02:53, Camaleón wrote:

(...)


¿Tienes algún disco USB de arranque instalado o alguna llave USB
conectada al equipo? No debería afectar a GRUB2.

Algun error he cometido. Efectivamente pude hacer funcionar supergrub2 y me
dice lo mismo que el grub en el disco y cuando le doy a Debian 10, se me
habre debian 11, yo quiero arrancar desde sda2 y no puedo, cuando voy a
gparted me doy cuenta que el sistema se ha montado en otras particiones (o
sea, las particiones que le puse en debian 11).

Qué raro :-?

Inicia Debian 11 normalmente y manda a la lista o sube a algún sitio
tipo pastebin (p. ej., «https://paste.debian.net/»), el contenido del
archivo de configuracuón de GRUB2 que tienes  en «/boot/grub/grub.cfg».

Recuerda que también puedes intentar iniciar tu Debian 10 desde la
línea de órdenes de GRUB2. Es un método un poco engorroso porque tendrás
que ir probando opciones pero funciona bien:

How to Rescue a Non-booting GRUB 2 on Linux
https://www.linux.com/training-tutorials/how-rescue-non-booting-grub-2-linux/

Saludos,


#
# DO NOT EDIT THIS FILE
#
# It is automatically generated by grub-mkconfig using templates
# from /etc/grub.d and settings from /etc/default/grub
#

(...)

Marcelo, el archivo que has enviado está incompleto (cortado), faltan
las líneas de arranque de los sistemas (menuentry).

Puedes enviarlo de nuevo a la lista (asegúrate de que copias/pegas
todas las líneas) o subir el archivo a algún sitio (como
https://www.file.io).

Saludos,


Que bobo, pasa que lo copié a mano. Que desastre.

Gracias por todo, ahí va el link

https://file.io/1buBSF9uwRXV 

Re: hello. Please help this big report go in the correct direction....

[Thomas Schmitt]

Hi,

(It seems that Timothy Danielson is not subscribed and missed the answers
  https://lists.debian.org/debian-user/2021/04/msg00410.html
  https://lists.debian.org/debian-user/2021/04/msg00411.html
  https://lists.debian.org/debian-user/2021/04/msg00412.html
)

In summary plus my own two cents:

The root of the problem is the apparent failure of
  sudo apt install google-drive-ocamlfuse
to install a command "google-drive-ocamlfuse" in reach of the shell.

The lengthy message after
  timdanielson@td546:~$ google-drive-ocamlfuse ~/mount/google-drive
stems from a helper program named "command-not-found" which gets triggered
by the shell because "google-drive-ocamlfuse" was not found.

Debian is said be in charge of the crash of "command-not-found".
It seems to be known since a while and has been worked on recently:
  https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=943989

But "google-drive-ocamlfuse" is not in Debian.

I would expect that the Ubuntu community is more able to debug why the
installation of google-drive-ocamlfuse from PPA failed.
Another approach would be to follow the advise at the bottom of
  https://github.com/astrada/google-drive-ocamlfuse
to open an issue with that problem.


Have a nice day :)

Thomas

Re: hello. Please help this big report go in the correct direction....

[Greg Wooledge]

On Wed, Apr 14, 2021 at 12:19:47PM +, Curt wrote:
> I was going to snidely inquire who in their right mind would create an
> executable called 'command-not-found', but then:
> 
> curty@einstein:~$ apt-cache show command-not-found
> 
> ...
> 
>  Description-en: Suggest installation of packages in interactive bash sessions
>   This package will install handler for command_not_found that looks up
>   programs not currently installed but available from the repositories.
> 
> Does this mean if you type 'blender' (a package available in the repositories)
> at the command line and don't happen to have blender installed,
> 'command-not-found' will somehow intervene to suggest installing it? (Question
> purely rhetorical (unless you'd want to answer it).)

Bash has a feature that permits the creation of a function named
command_not_found_handle.  If such a function exists, then instead of
giving the simple, traditional "bash: foobar: command not found"
message, bash will instead execute that function.

unicorn:~$ foobar
bash: foobar: command not found
unicorn:~$ command_not_found_handle() { echo "I'm a little teapot"; }
unicorn:~$ foobar
I'm a little teapot

Apparently the intent of this is to make the user's life better, but in
practice I'm not sure that's how it usually plays out.

I believe Ubuntu installs such a function in every user's bash sessions
by default, and uses it to try to tell the user which package they would
have to install to get the command they were trying to use.  In some cases,
I can see how that might be useful.  But sometimes, a command isn't found
because the user made a typo, or because the user has absolutely no clue
what they're doing, and then the results are just confusing.

Re: hello. Please help this big report go in the correct direction....

[Curt]

On 2021-04-14, Dan Ritter  wrote:
>
> This is not a problem with google-drive-ocamlfuse itself, though
> problems there might exist. This is a problem with a package
> literally called "command-not-found", which PureOS installed.
>

I was going to snidely inquire who in their right mind would create an
executable called 'command-not-found', but then:

curty@einstein:~$ apt-cache show command-not-found

...

 Description-en: Suggest installation of packages in interactive bash sessions
  This package will install handler for command_not_found that looks up
  programs not currently installed but available from the repositories.

Does this mean if you type 'blender' (a package available in the repositories)
at the command line and don't happen to have blender installed,
'command-not-found' will somehow intervene to suggest installing it? (Question
purely rhetorical (unless you'd want to answer it).)

Re: Mysql et php5.6

[Jean-Michel OLTRA]

Bonjour,


Le mercredi 14 avril 2021, ajh-valmer a écrit...


> Ce code ne fonctionne pas non plus.
> Il n'y a pas d'erreur.

C'est à dire ? La connexion ne se fait pas ? Tu n'as pas d'objet mysqli ?
Si ça échoue, que donne mysqli->connect_error (et connect_errno)

Regarde les propriétés de cet objet classe mysqli ici :
https://www.php.net/manual/fr/class.mysqli.php

> Mais mysqli me renvoie erreur.

Erreur, ou pas ? Avec connect_errno et connect_error, tu devrais en savoir
un peu plus sur l'erreur de connexion.

Sinon, tu as également les logs de mysql/mariadb, si tu y as accès.

A voir si ce ne serait pas un souci de connexion sur le socket unix sur
localhost vs la connexion tcp sur le port 3306 ?

-- 
jm

Re: SOLUCIONADO: Xerox B215 configurar scanner con TCP/IP en Debian 10 xfce

[Debian]

El 13/4/21 a las 19:58, Marcelo Giordano escribió:


On 12/4/21 09:14, Debian wrote:

El 12/4/21 a las 08:40, Debian escribió:


Pregunta sumamente estúpida:
En la configuración de la impresora, el protocolo IPP,¿Está activado? 
¿está abierto o bajo SSL/TSL?


Debería estar activado y abierto (sin seguridad SSL/TSL).

Esto lo podés ver desde un navegador web en http://IP-impresora.

Una vez asegurado que IPP (Internet Printing Protocol) esté 
funcionando, prueba con


$ scanimage -L

qué respuesta da. Tiene que ser algo parecido a esto (tengo una 
Samsung en esta red):
device `smfp:net;172.16.150.68' is a Samsung SCX-6x55 Series on 
172.16.150.68 Scanner
device `smfp:net;fe80::215:99ff:fe09:1afe%br0' is a Samsung SCX-6x55 
Series on fe80::215:99ff:fe09:1afe%br0 Scanner



JAP



Me olvidé de decirte que instales
# apt install sane-airscan

Luego, ejecutes

$ airscan-discover -t -d

JAP




Amigo:

Instalé Debian 11 kde y se solucionó automaticamente.

Gracias por tu ayuda.

Estoy en deuda



Me alegro.
Es como te dije en uno de los correos, a veces, los repositorios de 
Debian están más actualizados que los de Xerox.
El paso a seguir que te iba a decir era hacer una apt-pinning con los 
repositorios experimentales para cargar los controladores de cups.


Y gracias por cerrar el hilo y compartir.

JAP

Re: hello. Please help this big report go in the correct direction....

[Andrew M.A. Cater]

On Wed, Apr 14, 2021 at 06:38:18AM -0400, Dan Ritter wrote:
> Timothy Danielson wrote: 
> > don't read it if you feel huried,) I Respect what you do and I REALLY,
> > REALLY hate to ask for this level of assistance. I'm using PureOS, which I
> > guess is a derivative of Debian. I had started with 7.0 (Wheezy) and went
> > to 8.0. I never had the gonads to use Sid TBH. I ran the stables. Although
> > I ws a TINY variant by running a KDE desktop. I also had XFCE and LXDE?
> > installed, for really no other reason than I had read about them and
> > certain authors SWORE by them so I gave them a whirl.
> 
> 
> It appears that PureOS is a derivative of Debian. That means
> that the people responsible for PureOS changed Debian to suit
> their needs. Nobody outside of PureOS's maintainers and users is
> familiar with what they changed, so Debian users can only offer
> you limited help.
> 
> It looks like PureOS doesn't have a mailing list or forum for
> their users, only a wiki and a bug tracker at:
> 
> https://tracker.pureos.net/
> 
> In this case, there appear to be two issues:
> 
> > I messed up! there was a process that dely I umped more info and it
> > overwrote! ARRRGHH! I will try to reproduce it
> > 
> > https://www.maketecheasier.com/backup-files-to-google-drive-linux/
> > 
> > sudo add-apt-repository ppa:alessandro-strada/ppa
> 

Which distribution is this for? At a guess, Ubuntu since that's what is shown
on screen and _most_ PPAs have been created for Ubuntu.

Contact Alexander Strada is the only advice, I think, if it doesn't work as
a package. You run the risk of creating a FrankenDebian - a mixture from all
sorts of distributions which is almost always a problem in itself. In this 
instance, one package may be relatively easy to fix.

> 
> First issue: nobody except Alessandro Strada knows the contents
> of the packages in this repository. You should seek help from
> him.
> 
> > sudo apt update && sudo apt install google-drive-ocamlfuse
> > 
> > mkdir -p ~/mount/google-drive
> > 
> > I believe I created this dir and this is where the headaches started:
> 
> Creating that directory is unlikely to a problem in and of
> itself.
> 
> > timdanielson@td546:~$ google-drive-ocamlfuse ~/mount/google-drive
> > Could not find the database of available applications, run
> > update-command-not-found as root to fix this
> > Sorry, command-not-found has crashed! Please file a bug report at:
> > http://www.debian.org/Bugs/Reporting
> > Please include the following information with the report:
> > 
> > command-not-found version: 0.3
> > Python version: 3.7.3 final 0
> > Distributor ID: PureOS
> > Description:PureOS
> > Release:9.0
> > Codename:   amber
> 
> This is not a problem with google-drive-ocamlfuse itself, though
> problems there might exist. This is a problem with a package
> literally called "command-not-found", which PureOS installed.
> 
> You might fix this problem with
> 
> sudo update-command-not-found
> 
> or that might not help.

command-not-found also exists in Debian. It looks at the command you're trying
to run - if the package isn't installed, it offers to help you install the
package.

The command given above is the command that initialises command-not-found 
database so it is hightly likely that this will work to update an existing
database and fix the problem for you.

All the very best,

Andy C.

> 
> -dsr-
> 

Re: hello. Please help this big report go in the correct direction....

[Dan Ritter]

Timothy Danielson wrote: 
> don't read it if you feel huried,) I Respect what you do and I REALLY,
> REALLY hate to ask for this level of assistance. I'm using PureOS, which I
> guess is a derivative of Debian. I had started with 7.0 (Wheezy) and went
> to 8.0. I never had the gonads to use Sid TBH. I ran the stables. Although
> I ws a TINY variant by running a KDE desktop. I also had XFCE and LXDE?
> installed, for really no other reason than I had read about them and
> certain authors SWORE by them so I gave them a whirl.


It appears that PureOS is a derivative of Debian. That means
that the people responsible for PureOS changed Debian to suit
their needs. Nobody outside of PureOS's maintainers and users is
familiar with what they changed, so Debian users can only offer
you limited help.

It looks like PureOS doesn't have a mailing list or forum for
their users, only a wiki and a bug tracker at:

https://tracker.pureos.net/

In this case, there appear to be two issues:

> I messed up! there was a process that dely I umped more info and it
> overwrote! ARRRGHH! I will try to reproduce it
> 
> https://www.maketecheasier.com/backup-files-to-google-drive-linux/
> 
> sudo add-apt-repository ppa:alessandro-strada/ppa


First issue: nobody except Alessandro Strada knows the contents
of the packages in this repository. You should seek help from
him.

> sudo apt update && sudo apt install google-drive-ocamlfuse
> 
> mkdir -p ~/mount/google-drive
> 
> I believe I created this dir and this is where the headaches started:

Creating that directory is unlikely to a problem in and of
itself.

> timdanielson@td546:~$ google-drive-ocamlfuse ~/mount/google-drive
> Could not find the database of available applications, run
> update-command-not-found as root to fix this
> Sorry, command-not-found has crashed! Please file a bug report at:
> http://www.debian.org/Bugs/Reporting
> Please include the following information with the report:
> 
> command-not-found version: 0.3
> Python version: 3.7.3 final 0
> Distributor ID: PureOS
> Description:PureOS
> Release:9.0
> Codename:   amber

This is not a problem with google-drive-ocamlfuse itself, though
problems there might exist. This is a problem with a package
literally called "command-not-found", which PureOS installed.

You might fix this problem with

sudo update-command-not-found

or that might not help.

-dsr-

Re: Mysql et php5.6

[ajh-valmer]

On Tuesday 13 April 2021 00:06:27 Jean-Michel OLTRA wrote:
> Le lundi 12 avril 2021, ajh-valmer a écrit...
> > > > if (!$CNX = mysql_connect($DB_BASE, $DB_USER, $DB_PASS, $DB_SERVER)) {
> 
> > La syntaxe ci-dessus n'est pas supprimée en php 5.6.
> > Normalement, il devrait accepter les scripts php 5

> Que mysql_connect existe encore c'est une chose.
> Que l'ordre des paramètres que tu passes dans ce que tu as écrit soit
> valide, ça en est une autre. A moins que tu aies mal recopié ton véritable
> code de connexion ?
> Ceci étant, je n'utilise pas mysql_connect, 
> mais "new mysqli(server, username, password, db)".

Ce code ne fonctionne pas non plus.
Il n'y a pas d'erreur.

J'ai lu qu'il fallait utiliser mysqli ou PDO à partir de la version 7.
Mais mysqli me renvoie erreur.

Bonne journée.

Re: Efi et Debian Buster

[Georges]

Le Wed, 14 Apr 2021 09:36:40 +0200,

PM a écrit :

> Merci pour tous ces retours d’expérience.
> 
> Ceci état je m’étonne toujours de la découverte actuelle de GPT… Il
> me semblait depuis longtemps que MSDOS était abandonné.
> 
> Une question en passant : est-ce que tous ces problèmes venaient du
> microprogramme installé (BIOS) ou d’autre chose ?

 J'ai bricolé des quantité de BIOS depuis 198? sur des tours, des
 portables, mais celui de ce nouveau PC c'est l’horreur. Même pas foutu
 de lui demander de botter 1 sur CD 2 sur clef USB 3 sur HD.  
> 
> Encore merci
> 
> > Le 14 avr. 2021 à 08:28, Georges  a écrit :
> > 
> > 
> > Le Fri, 9 Apr 2021 21:24:10 +0200,
> > 
> > G a écrit :
> > 
> >> Bonsoir à tous,
> >> 
> >> J'ai apprécié toutes les réponses. Je teste tous cela ce soir et
> >> demain.
> >> 
> >> Je vais essayer Bullseye qui était installé sur l'ancien et
> >> marchait très bien.
> >> 
> >> Bonne nuit et je reviens ...
> > 
> > Me revoilà,
> > 
> > Maintenant  UEFI / EFI  trituré par MS.
> > Sur un ordinateur récent, il ma fallut partitioner le disque en GPT
> > ( gparted en live). Installer Bullseye net install et avec un CD
> > rEFInd contourner le BIOS pour enfin accéder au bon grub (merci au
> > génial programmeur). Sa ma pris 4 jours pour comprendre tout ce
> > bazar. Merci MS.
> > 
> >  Mes partitions !
> > 
> > GPT autorise plus de partitions "primaire"
> > 
> > sda ;
> > 1  /  Bullseye,  2 /Var  Bullseye,  3 /home Bullseye,
> > 4  / Buster (permet de dépanner Bullseye et de la sauver avec
> > "rsync"), 5  /home Buster,
> > 6  pour tester une autre distribution,
> > 7  /  Sid pour voir l'évolution et les nouveautées de Debian,
> > 8  tous MesFich perso,
> > 9  tous MesFExtension,
> > 10  une réserve pour pour étendre au cas ou,
> > 11   SWAP 2,94 Go,
> > 12   Fat32 pour EFI 699,00  Mo  qui sera montée dans /boot/efi par
> > fstab
> > 
> ># /mnt/sda12 was on /boot/efi   perso
> > UUID=E795-39FC /boot/efi vfat  defaults  02
> >#
> > je l'ai inventé mais si ce n'est pas correct corigez moi
> > 
> > Merci de votre attention
> > 
> > 
> 
> --
> Pierre Malard
> 
>   « on ne risque rien à livrer le secret professionnel car
>  on ne livre pas la façon de s’en servir »
>   Jean Cocteau - « Le secret professionnel »
> - 1922
> 
>|\  _,,,---,,_
>/,`.-'`'-.  ;-;;,_
>   |,4-  ) )-,_. ,\ (  `'-'
>  '---''(_/--'  `-'\_)   πr
> 
> perl -e '$_=q#: 3|\ 5_,3-3,2_: 3/,`.'"'"'`'"'"' 5-.  ;-;;,_:  |,A-  )
> )-,_. ,\ (  `'"'"'-'"'"': '"'"'-3'"'"'2(_/--'"'"'  `-'"'"'\_):
> 24πr::#;y#:#\n#;s#(\D)(\d+)#$1x$2#ge;print'
> - --> Ce message n’engage que son auteur <--
> 

hello. Please help this big report go in the correct direction....

[Timothy Danielson]

Hello,

I apologize for my  mental deficits. (I will leave that story for the end,
don't read it if you feel huried,) I Respect what you do and I REALLY,
REALLY hate to ask for this level of assistance. I'm using PureOS, which I
guess is a derivative of Debian. I had started with 7.0 (Wheezy) and went
to 8.0. I never had the gonads to use Sid TBH. I ran the stables. Although
I ws a TINY variant by running a KDE desktop. I also had XFCE and LXDE?
installed, for really no other reason than I had read about them and
certain authors SWORE by them so I gave them a whirl.

O.K. as i'm sure you will see, I am operating PureOS. I haven't honestly
kept up as well as I would wish as my internet is EXCLUSIVELY through my
cell phone now. (What can I say, it's one of the "wonders" of being poor in
the USA, which is better than poor in any 3rd world country. I'm now
complaining, simply sharing that I'm trying to do my best with what I
have.) Again I digress (OK, I'm 90% digressed! I wY to be brief with my
story and after that will be a cut and paste of EVERYTHING I xould put on
the screen so you could see what U saw and hope it will give you some
insight into what is going wrong and how t fix the problem. At this point,
I am siply going to hopr yhat this is not 100% the result of my failed
updating.

*PERSONAL STORY*: Ok, do NOT read this if pressed for time, however, you
may possibly, in some way  benefit from reading "some other blokes" rough
story. I recieved my TBI (Traumatic Brain Injury) On Februry 23rd, 2006.
How do I know this? well, when I was in coma; (from what I experienced was
on the average for those that were able to resume life afterwards;) I had
lost orientation to where and WHEN I was! I had some misunderstandings. I
THOUGHT I was a dew YEARS earlier than when I was and did not know I was
HOME... NOT in the ARMY! So, when I was awoken at 9:00 a.m. and saw the sun
FULLY RISEN, shining in it's blazing glory, not like it i at o'dark thirty
in the morning, i was freaked out! I started getting frustrated at some
poor nurse, (which I later felt SO BAD about my mis-reaction that I made a
trip JUST to apologize to her!) I yelled something to the effect of, "Why
did you wake me for PT so LATE!!! The 1SG (First Sargent) is going to
f*ck!n* kill me! I'll be article 15'ed! I don't want that $hi! (it was
understood in the military article 15 is 1/2 pay + extra duty for 2 weeks +
deferred promotion (which is MORE lost pay,))  (suffice to say, was REALLY,
REALLY, bad! Thuse, was avoided TREMENDOUSLY!)   poor, poor nurse I
wsn't 100% swering 2 of every 3 words forming a constructive and coherent
sentence... yet. BUT she was able to glimse down that hall nd see the room
on the end.


Long story short again, that was after my INITIAL brain injury which
honestly was HUGE for me! It was like taking a 52 card deck + 2 jokers and
then taking the jokers out and 5 random cards which NOBODY knew WHICH
specific ones were taken out. After that I had a couple shunt "revisions",
which involve me being put to sleep and have them open my head to repair
the "shunt", a device that basically helps my body do what most do
naturally. I wke up 90-95% of who I was before, much like the initial TBI.
but 95% of 95%f 95% is GREAT! wgn looking at the alternative OF COURSE! but
95% of 95%f 95% is 85.7%... not terrible but that is on the upper end too,
and a 4th time will be 85.4%

/ if it were even slightly less like 92.5% it would be 79.2% ans 73.2%
respectively. My mistake was being god at math and NOT takinf it as a
major! DOH!


as promised, here us *EVERYTHING* from my *terminal window*:


I messed up! there was a process that dely I umped more info and it
overwrote! ARRRGHH! I will try to reproduce it

https://www.maketecheasier.com/backup-files-to-google-drive-linux/

sudo add-apt-repository ppa:alessandro-strada/ppa

sudo apt update && sudo apt install google-drive-ocamlfuse

mkdir -p ~/mount/google-drive

I believe I created this dir and this is where the headaches started:

google-drive-ocamlfuse ~/mount/google-drive

strangely I had a similar readout after I attempted to run
reportbug but i am honeslty a bit foggy on what I did:

timdanielson@td546:~$ google-drive-ocamlfuse ~/mount/google-drive
Could not find the database of available applications, run
update-command-not-found as root to fix this
Sorry, command-not-found has crashed! Please file a bug report at:
http://www.debian.org/Bugs/Reporting
Please include the following information with the report:

command-not-found version: 0.3
Python version: 3.7.3 final 0
Distributor ID: PureOS
Description:PureOS
Release:9.0
Codename:   amber
Exception information:

local variable 'cnf' referenced before assignment
Traceback (most recent call last):
  File "/usr/share/command-not-found/CommandNotFound/util.py", line
23, in crash_guard
callback()
  File "/usr/lib/command-not-found", line 93, in main
if not cnf.advise(args[0], 

Re: Efi et Debian Buster

[Eric Degenetais]

Le mer. 14 avr. 2021 à 08:57, Michel  a écrit :
>
> Le 14/04/2021 à 08:30, Georges a écrit :
> >
> > Le Fri, 9 Apr 2021 21:24:10 +0200,
> >
> > G a écrit :
> >
> >> Bonsoir à tous,
> >>
> >>  J'ai apprécié toutes les réponses. Je teste tous cela ce soir et
> >>  demain.
> >>
> >>  Je vais essayer Bullseye qui était installé sur l'ancien et marchait
> >>  très bien.
> >>
> >>  Bonne nuit et je reviens ...
> >
> >  Me revoilà,
> >
> > Maintenant  UEFI / EFI  trituré par MS.
> > Sur un ordinateur récent, il ma fallut partitioner le disque en GPT (
> > gparted en live). Installer Bullseye net install et avec un CD rEFInd
> > contourner le BIOS pour enfin accéder au bon grub (merci au génial
> > programmeur). Sa ma pris 4 jours pour comprendre tout ce bazar.
> >   Merci MS.
> >
>
> Tu as acheté un portable sans OS, qu'est-ce MS a à voir là dedans?


Avoir mis ses doigts sales dans la conception d'UEFI et du secure
boot, avec pour effet d'em**er toute personne qui veut autre chose que
windows sur un PC avec UEFI, je pense...

Cordialement

Éric Dégenètais

Re: Captura de pantalla (video screencast) con FFMPEG con muy bajos FPS y poca fluidez

[Eduardo Jorge Gil Michelena]

 Mira...
Yo tengo una empresa que trabaja editando videos publicitarios en forma 
profesional para Cine, TV y Streaming así que de video tenemos bastante 
experiencia.
Es por ello que te puedo asegurar que Linux está tremendamente atrasado en todo 
lo que sea video y gráfica profesional por lo que no me sorprende que al 
actualizar una maquina a una muchísimo más moderna el comportamiento en el 
trabajo de video sea menos eficiente y más lento. A nosotros nos pasa a menudo, 
con máquinas de 16 núcleos Xeon la sola transposición para sólo adaptar las 
frames a una norma, es una tortura pues es tremendamente lenta sobre Linux. 

Una solución de compromiso que no resuelve totalmente los tremendos atrasos de 
Linux en cuestión de videos es actualizar los drivers e instalar MESA. Eso, con 
suerte y viento a favor, te podrá sacar momentáneamente del paso, pero no te 
solucionará el tema de la lentitud que tiene LINUX al trabajar video y muchos 
más con hardware gráfico nuevo.

Nosotros hace rato que DEJAMOS de USAR Linux para trabajos en video y gráfica 
porque NO sirve, esta tan desactualizado que los empleados editores se quejaban 
porque les hacía perder mucho tiempo y los programas que tenían a mano para 
editar eran muchísimo más pobres que los que puede haber en Windows o Macs. Así 
que dejamos de usar Linux en gráfica, usamos WINDOWS para gráfica y videos 
mientras dejamos LINUX en servers y PC Administrativas. 

Para finalizar: Con Linux y algunos programitas se pueden hacer videos pero de 
manera NO profesional y con un desperdicio de recursos impresionante. Conocemos 
que hay películas, animaciones y clips hechos enteramente con LINUX pero... lo 
que no dicen es que lo hecho llevó muchísimo más tiempo, más recursos de PC y 
mucho más esfuerzo que si lo hubiesen realizado con Windows. 
 El lunes, 12 de abril de 2021 16:23:25 ART, jEsuSdA 8) 
 escribió:  
 
 Hola, compañeros.

Recientemente he cambiado de PC, pasé de tener un Intel Core i5 4460 con 
tarjeta gráfica integrada a un Xeon E5 2678 v3 con gráfica AMD RADEON RX 
550.

Sobre el papel, el nuevo PC es del orden de 3 a 7 veces más potente que 
el antiguo y puedo atestiguar que así es en el uso diario, edición de 
vídeo e imagen etc. La ventaja de tener tantos núcleos e hilos 
disponibles es palplable. A nivel juegos no lo he probado porque 
realmente no soy gamer y los pocos juegos que uso son los típicos libres 
que vienen con Debian y algún emulador que, sinceramente, ya funcionaban 
bien con el PC antiguo.

Sin embargo hay una tarea que me trae de cabeza por su terrible 
desempeño: la captura de pantalla en vídeo.

Con mi antiguo PC era capaz de capturar a más de 60 fps a pantalla 
completa mientras realizaba cualquier tarea de la que necesitara hacer 
videotutorial.

Incluso con mi lenovo thinkpad x230 logro capturar pantalla a más de 
80fps con total fluidez.

El comando que he empleado siempre es:

ffmpeg  -f x11grab -draw_mouse 1 -framerate 60 -video_size 1920x1200 -i 
:0.0+1680,0  -qscale 0 -pix_fmt yuv420p -c:v libx264 -preset medium -qp 
0 -q:v 1 -s 1920x1200 -f matroska -threads 4 video.mkv

Notas:

-video_size 1920x1200 -i :0.0+1680,0 y -s 1920x1200 son las dimensiones 
y posición de la región a capturar (mi monitor derecho).

Nótese que incluso usaba -preset medium y codificación por software, de 
forma que obtenía muy buena calidad incluso con esa configuración de 
parámetros y sin bajar nunca de los 60 fps.



¿Qué me ocurre ahora?

El equipo es incapaz de capturar a más de 20 fps lo que hace que 
cualquier vídeo sea inválido al ir a saltos y no llegar siquiera a 
30fps, que sería lo mínimo exigible.

Además, se nota bastante la merma de respuesta del equipo en cuanto 
lanzo el comando. Es decir, toda esa fluidez y suavidad que se aprecia 
cuando se trabaja normalmente, desaparece y hasta el desplazar una 
ventana de un lado a otro resulta tosco y a trompicones.


He probado con distintos parámetros de ffmpeg, a capturar en bruto, sin 
codificar.

He probado a guardar el vídeo resultante directamente en un disco en 
RAM, para evitar el posible cuello de botella de la escritura en disco. 
No afecta en absoluto.




Así las cosas, ¿Alguien tiene alguna sugerencia de, al menos, por donde 
puedo seguir indagando para hallar solución al problema?





Datos adicionales, por si son de ayuda:

$ → inxi
CPU: 12-Core Intel Xeon E5-2678 v3 (-MT MCP-)
speed/min/max: 1201/1200/3300 MHz Kernel: 5.10.0-0.bpo.4-amd64 x86_64
Up: 1d 6h 55m Mem: 6427.6/32012.4 MiB (20.1%)
Storage: 13.76 TiB (55.9% used) Procs: 433 Shell: bash 5.0.18 inxi: 3.0.32



$ → ffmpeg -v
ffmpeg version 4.1.6 Copyright (c) 2000-2020 the FFmpeg developers
   built with gcc 8 (Debian 8.3.0-6)
   configuration: --disable-decoder=amrnb --disable-decoder=libopenjpeg 
--disable-libopencv --disable-outdev=sdl2 --disable-podpages 
--disable-sndio --disable-stripping --enable-libaom --enable-avfilter 
--enable-avresample --enable-gcrypt --disable-gnutls --enable-openssl 

Re: Efi et Debian Buster

[Pierre Malard]

Merci pour tous ces retours d’expérience.

Ceci état je m’étonne toujours de la découverte actuelle de GPT… Il me semblait 
depuis longtemps que MSDOS était abandonné.

Une question en passant : est-ce que tous ces problèmes venaient du 
microprogramme installé (BIOS) ou d’autre chose ?

Encore merci

> Le 14 avr. 2021 à 08:28, Georges  a écrit :
> 
> 
> Le Fri, 9 Apr 2021 21:24:10 +0200,
> 
> G a écrit :
> 
>> Bonsoir à tous,
>> 
>> J'ai apprécié toutes les réponses. Je teste tous cela ce soir et
>> demain.
>> 
>> Je vais essayer Bullseye qui était installé sur l'ancien et marchait
>> très bien.
>> 
>> Bonne nuit et je reviens ...
> 
> Me revoilà,
> 
> Maintenant  UEFI / EFI  trituré par MS.
> Sur un ordinateur récent, il ma fallut partitioner le disque en GPT (
> gparted en live). Installer Bullseye net install et avec un CD rEFInd
> contourner le BIOS pour enfin accéder au bon grub (merci au génial
> programmeur). Sa ma pris 4 jours pour comprendre tout ce bazar.
>   Merci MS.
> 
>  Mes partitions !
> 
> GPT autorise plus de partitions "primaire"
> 
> sda ;
> 1  /  Bullseye,  2 /Var  Bullseye,  3 /home Bullseye,
> 4  / Buster (permet de dépanner Bullseye et de la sauver avec "rsync"),
> 5  /home Buster,
> 6  pour tester une autre distribution,
> 7  /  Sid pour voir l'évolution et les nouveautées de Debian,
> 8  tous MesFich perso,
> 9  tous MesFExtension,
> 10  une réserve pour pour étendre au cas ou,
> 11   SWAP 2,94 Go,
> 12   Fat32 pour EFI 699,00  Mo  qui sera montée dans /boot/efi par fstab
> 
># /mnt/sda12 was on /boot/efi   perso
> UUID=E795-39FC /boot/efi vfat  defaults  02
>#
> je l'ai inventé mais si ce n'est pas correct corigez moi
> 
> Merci de votre attention
> 
> 

--
Pierre Malard

  « on ne risque rien à livrer le secret professionnel car
 on ne livre pas la façon de s’en servir »
  Jean Cocteau - « Le secret professionnel » - 1922

   |\  _,,,---,,_
   /,`.-'`'-.  ;-;;,_
  |,4-  ) )-,_. ,\ (  `'-'
 '---''(_/--'  `-'\_)   πr

perl -e '$_=q#: 3|\ 5_,3-3,2_: 3/,`.'"'"'`'"'"' 5-.  ;-;;,_:  |,A-  ) )-,_. ,\ 
(  `'"'"'-'"'"': '"'"'-3'"'"'2(_/--'"'"'  `-'"'"'\_): 
24πr::#;y#:#\n#;s#(\D)(\d+)#$1x$2#ge;print'
- --> Ce message n’engage que son auteur <--



signature.asc
Description: Message signed with OpenPGP

Re: Efi et Debian Buster

[Michel]

Le 14/04/2021 à 08:30, Georges a écrit :
> 
> Le Fri, 9 Apr 2021 21:24:10 +0200,
> 
> G a écrit :
> 
>> Bonsoir à tous,
>>
>>  J'ai apprécié toutes les réponses. Je teste tous cela ce soir et
>>  demain.
>>
>>  Je vais essayer Bullseye qui était installé sur l'ancien et marchait
>>  très bien.
>>
>>  Bonne nuit et je reviens ...
> 
>  Me revoilà,
> 
> Maintenant  UEFI / EFI  trituré par MS.
> Sur un ordinateur récent, il ma fallut partitioner le disque en GPT (
> gparted en live). Installer Bullseye net install et avec un CD rEFInd
> contourner le BIOS pour enfin accéder au bon grub (merci au génial
> programmeur). Sa ma pris 4 jours pour comprendre tout ce bazar. 
>   Merci MS.
> 

Tu as acheté un portable sans OS, qu'est-ce MS a à voir là dedans?

Re: Efi et Debian Buster

[Georges]

Le Fri, 9 Apr 2021 21:24:10 +0200,

G a écrit :

> Bonsoir à tous,
> 
>  J'ai apprécié toutes les réponses. Je teste tous cela ce soir et
>  demain.
> 
>  Je vais essayer Bullseye qui était installé sur l'ancien et marchait
>  très bien.
> 
>  Bonne nuit et je reviens ...

 Me revoilà,

Maintenant  UEFI / EFI  trituré par MS.
Sur un ordinateur récent, il ma fallut partitioner le disque en GPT (
gparted en live). Installer Bullseye net install et avec un CD rEFInd
contourner le BIOS pour enfin accéder au bon grub (merci au génial
programmeur). Sa ma pris 4 jours pour comprendre tout ce bazar. 
Merci MS.

  Mes partitions !

GPT autorise plus de partitions "primaire"

sda ; 
1  /  Bullseye,  2 /Var  Bullseye,  3 /home Bullseye,  
4  / Buster (permet de dépanner Bullseye et de la sauver avec "rsync"),  
5  /home Buster, 
6  pour tester une autre distribution,  
7  /  Sid pour voir l'évolution et les nouveautées de Debian,  
8  tous MesFich perso,
9  tous MesFExtension, 
10  une réserve pour pour étendre au cas ou,  
11   SWAP 2,94 Go,  
12   Fat32 pour EFI 699,00  Mo  qui sera montée dans /boot/efi par fstab

# /mnt/sda12 was on /boot/efi   perso
 UUID=E795-39FC /boot/efi vfat  defaults  02
#
je l'ai inventé mais si ce n'est pas correct corigez moi

Merci de votre attention  
 

Re: TDE Bug: what package responsible

[deloptes]

Dan Ritter wrote:

> q4os appears to be a Debian derivative, so presumably they would
> know better than we as to what is going on. I can't find a
> mailing list, but I do see a forum at
> https://www.q4os.org/forum/
> 
> I'm going to guess that they will want a better explanation of
> the symptoms than "tdestartupconfig complains and goes tits-up".

yes and also Q4OS would probably not solve the problem as TDE has its own
user mailing list.

Please, do not forget to attach what exactly tdestartupconfig complains for.